It began, you could say, with the Y2K scare. Fearful that their computers would come crashing down on New Year's Day, companies in the late 1990s scrambled to embrace massive new software packages, particularly enterprise resource planning (ERP) systems. And while Y2K proved to be something of a non- event, many of those businesses ended up in a different kind of nightmare.

Call it the result of promotional hype by software vendors and the media, combined with a looming deadline. But a number of companies took the ERP leap without first peering over the edge of the cliff. They were surprised when implementations stretched into months, in some cases years, with costs running into the millions.

Most of all, they underestimated the degree to which their business processes and organizational structures had to change. ERP systems demand centralized databases and the easy flow of information between departments that may barely be on speaking terms. They call for a housecleaning of legacy systems that may have grown up haphazardly and are incompatible with one another. For many newcomers to ERP, a single piece of software meant a radically new way of doing business.

Those same companies may be a bit gun-shy about taking the next logical step in the acquisition of information technology: supply-chain management, planning and execution systems. In fact, just about any major software package that promises to automate business systems raises the same uneasy questions: How will this product change my core processes? And which comes first, business-process reengineering or the software that necessitates it?

Experts give a variety of answers. But all point to the deeply intertwined nature of process and technology in the age of electronic commerce and the internet. The search for IT excellence, they say, involves a complex series of steps, with each element feeding on the other.

Burned by what they see as the tyranny of ERP, supply-chain executives on the reengineering path might be tempted to leave their choice of software until the very end. They shouldn't. While vendors should never be allowed to dictate the terms of an organizational restructuring, they have a valuable role to play throughout the process.

They may even spark a new practice by offering technology that didn't previously exist. Witness the explosion of trading exchanges and internet-based auctioning over the past year. "I've always liked the idea of software as a catalyst for reengineering," says Barry Wilderman, vice president of MetaGroup in Stamford, Conn.

He calls for "intelligent reengineering," in which software packages for supply-chain planning and related functions serve as guides to the process. He urges that a company's IT experts be brought into the picture early, right alongside the business team, to ensure a smooth transition.

Often a software vendor can make suggestions for improvements, or identify areas of need, that hadn't occurred to the prospective customer. "If you go down the road of process definition in excruciating detail, you limit the creativity of different vendors considerably," says Gary Layton, vice president of marketing with InterBiz, a division of Islandia, New York-based Computer Associates.

First, a Strategy
The actual choice of software vendor doesn't come at the start, however. First, executives need to define their broad objectives and business requirements, says Michael Lipton, director of supply-chain consulting with SAP America in Foster City, Calif. The ultimate solution must support the desired business processes, not the other way around.
Mark Wheeler, a national principal of the supply-chain and logistics practice of IBM Global Services in New York City, puts his clients through a rigorous analysis of business strategy. Frequently the project will be spurred by a specific problem, such as excess inventory or a congested loading dock.

But the answer tends to lie deeper than that. It could require a close look at inventory management, advance procurement or vendor compliance -- not just the acquisition of off-the-shelf warehouse management system (WMS) software. Says Wheeler: "Going beyond the warehouse is almost always part of the solution."

The right approach to reengineering depends entirely on context, says Al Delattre of Andersen Consulting in Los Angeles. It is determined by where a company adds value in the supply chain, which capabilities it needs to do the job, and the degree to which it outsources key functions. Some businesses, for example, are heavily oriented toward physical processes, and need technology to make them easier. Others are more systems-savvy and need only slight changes in the way they do things. A similar gap exists between "bleeding-edge" companies and those that prefer to follow the innovators.

Every company needs a clear vision of the way it wants to be known within its industry, says Kevin Coleman, chief strategist of the iPlanet/Netscape division of AOL in Bethesda, Md. Citing a 1995 book by Michael Treacy, "The Discipline of Market Leaders" (Perseus Publishing), he argues that there are three distinct ways of going to market: through customer intimacy, operational excellence, and innovation.

No company can possibly be all three at once, Coleman says. The decision to achieve customer intimacy may drive up cost and threaten operational excellence, which is associated with lean supply chains and low prices. A company's choice of category -- its "branding," Coleman calls it -- is essential to determining how it will reengineer business processes, and which software it will use for the task.

A Call for Change
At the dawn of the internet age, all companies share at least one imperative: the need for change. Those who have failed to launch a full-scale strategy review may be unaware of the price of maintaining the status quo. Wheeler's group at IBM defines a client's current practices, then projects the likely cost of continuing them into the future. "It's a real eye- opener," he says. "Often it results in real interesting discussions between sales and marketing and operations."

Having gathered their core requirements, companies can next move to prototyping new business processes. At this stage, they compare their plans with what's available in the marketplace, as well as flush out additional requirements based on feedback from vendors.
They may be forced to revise those plans due to the limitations of available software. But any changes made for that reason should amount to little more than "a tweak," says Lipton. Most of all, they should avoid allowing any system, especially in the area of supply-chain planning, to wrest the decision-making process from humans.

"It's not healthy to treat software as a black box," Lipton says. "You need to be able to intervene if you have other requirements."

The acquisition of an ERP package doesn't perfectly prepare a company to reengineer its supply-chain planning function. For one thing, the former is a more massive job, spanning an entire organization. Supply-chain planning (SCP) issues typically involve less than 20 internal users, says Jeff Schutt, a partner with CSC Consulting in Austin, Tex. And SCP packages, consisting of a number of point solutions, aren't likely to cause similar upheavals in the organization.

The primary change that arises from implementation of SCP software is the consolidation of capacity management, says Russ King, director of industry operations with Rockville, Md.-based Manugistics. Users quickly learn of the benefits to be derived from centralized control of inventory, if not the actual physical goods. "Otherwise," says King, "organizational changes are not that severe, compared with ERP."

The two kinds of software nevertheless share some basic practices on the part of their users. Schutt says companies have tended to develop a precise set of software requirements before selecting a vendor. They might spend large amounts of time and money on customizing the software to their supposedly unique needs.

A Cloudy Vision
But that approach has a flaw: It assumes that executives have a clear vision from the start of what their business processes ought to be. Most organizations have limited knowledge of a given industry's leading-edge processes, says Schutt. Instead, they rely on software providers -- hardly the most objective parties -- for information on benchmarking.

"Some software products force adherence to vendor-defined best practices," says Layton. "It paints you pretty solidly into a corner."

A better way is to start with the overall vision, then bring in packaged-software vendors at a relatively early stage, says Schutt. In this way, providers can be subjected to on-site testing. CSC's Solution Demonstration Lab matches a company's detailed list of requirements with the capabilities of a particular program. In many cases the pilot can be undertaken before a company has paid the vendor's substantial licensing fee.

The consultancy of Pattiglio, Rabin Todd & McGrath similarly urges the use of a "conference-room pilot" to design and simulate the performance of supply-chain planning and execution software. Such an approach is "essential to the identification of process and technology gaps that will need to be filled to enable a smooth transition to the new processes and systems," say PRTM analysts Todd Ackerman and Mark Stonich in a 1999 paper.

Getting the vendor to agree to a pilot before payment of the license fee isn't always easy. Schutt acknowledges that providers aren't quick to agree to that requirement, which entails a substantial investment on their part. But most will go along, especially if the client is a large and lucrative customer.

Vendor selection is an ability unto itself. King says Fortune 500 companies are beginning to do a better job of up-front process design, followed by fine-tuning with the help of prospective IT systems. Many have learned from the experience of ERP, where the choice of vendor wasn't always preceded by a lot of internal soul-searching.

To most vendors today, it's not just about the source code. They shy away from the label of mere software sellers. Manugistics views itself as a purveyor of business processes, the software solutions that support them, and the architecture that makes them work, says King. It's the combination of the three that brings benefits to a customer.

True success in reengineering comes from the ability to juggle all elements -- strategy, process and software -- in a virtually simultaneous manner. Wheeler speaks of "concurrent design integration," whereby business managers work side by side with technology experts to craft a solution. Like the modern-day supply chain itself, business process reengineering is anything but a linear process.

The Need to Collaborate
As if the need for wholesale internal restructuring weren't enough, companies are faced with an additional requirement: the need to collaborate with channel partners. ERP was designed to revamp a user's internal operations. Surprisingly, so were most supply-chain management programs.

The term "supply chain" is supposed to connote a whole universe of businesses working together to serve a common customer. But SCP packages are crafted primarily for big internal systems, claims Schutt. Only recently have vendors begun exploring ways to get outside the walls of an organization, with software that allows various systems to "talk" to one another, and data to be viewed and acted on by multiple partners.

The hot topic is collaboration, and it's forcing companies to take a whole new approach to business-process design. "The question is, 'What is your collaborative strategy?'" says Bob Moncrieff, a director of PRTM in Mountain View, Calif. Companies must decide which are their core competencies, and which can be handed over to outside providers. The answer will determine the applicable software and the degree of customization required.

Current examples of collaboration include the Collaborative Planning, Forecasting and Replenishment (CPFR) model, under development within the retail and consumer goods industries. After five years, CPFR is only just getting off the ground. The delay, says Lipton, was caused in part by the unwillingness -- or inability -- of companies to exchange critical data and construct joint business plans.

Also lagging until recently was the software to realize those goals. Now a large number of vendors are offering modules that support CPFR. The human factor, requiring a brand new attitude by top management toward supply-chain business processes, has been slower to come along. Unanswered questions include the source of a joint forecast, who's responsible for it, and how it can be reconciled with the budgets of all concerned.

The favorite means of collaboration today is, of course, the internet. With its ease of access and low cost, the technology is fast surpassing the much-touted alternative of electronic data interchange (EDI). Yet the net, too, is in its infancy, and few companies grasp the impact it will have on business processes in the decade to come.

One thing it has already done is highlight, sometimes to an embarrassing degree, corporate shortcomings. Nothing looks worse than a half-baked strategy to sell or procure via the World Wide Web. Take the spectacular failures of big name retailers to service customers over the past two Christmas seasons. Says Delattre: "The internet will exacerbate business problems at a much greater rate and amplitude than we've ever seen before."

The technology's most obvious quality is speed. Companies looking to exploit the medium -- and that's just about everyone --are having to rewrite business plans to reflect their newfound ability to connect with customers and vendors in real time, or something close to it. Meanwhile, buyers on the web are becoming increasingly impatient.

What's It All Mean?
The far-reaching implications are harder to predict. Even now, Moncrieff suggests, companies are being too timid in their approach to the internet. Understandably, they are focused on the need for standards, such as RosettaNet messaging in the high-tech industry. What they're missing is the chance to challenge business norms. They have yet to achieve the kind of critical mass that will create brand new marketplaces and the real-time flow of data among channel partners. According to Coleman, less than 2 percent of world commerce is conducted over the internet today. That number will jump to between 20 and 25 percent by 2010.

For now, most executives are still wrestling with the challenge of implementing basic software. In the process, they face yet another key decision: to customize or not? No off-the-shelf package can meet the precise needs of every user; some degree of tinkering is always required. But customization is costly, and it creates severe difficulties when a new version of underlying software is released.

Moncrieff advises clients to keep customization to a minimum. Work on only those pieces of the system that are strategically important, he says. Elsewhere, it might be wise to accept a vendor's "vanilla" offering, even if some compromises are required.

For mid-market companies, who have limited budgets and smaller IT staffs, customization is even more of a no-no. When the software doesn't match the business process, says Layton, "reengineer the business process."

King draws a distinction between customization and configuration. In the first instance, new code must be written. In the second, a software package is equipped with numerous options for processes such as replenishment, forecasting and customer orders. Companies can tailor the system to their own needs.

The whole question of customization grows less critical as software becomes more configurable and user-friendly. Coleman calls the original ERP systems "dinosaurs," with their monolithic structures, complex business rules and client servers. Today, the internet allows users access to business software in a hosted mode, requiring only a browser and a central server that is somebody else's problem. Updates are easier to install, and software solutions more targeted. As a result, users are free to pursue a best-of-breed strategy, choosing from multiple vendors instead of a single provider that can dictate a user's business processes.

Still, the true marriage of process and technology, driven by the ultimate promise of e-commerce and the internet, has yet to occur. Companies continue to search for the business plan and supporting software that will ensure competitive advantage, even as the rules change daily. Most remain in the experimental stage.

Moncrieff likens the situation to sailboats jockeying for position at the starting line of a race, just before the gun goes off. "Once somebody figures out what it means," he says, "it's going to expand very quickly."